THIRD CIRCUIT FAVORS AGGREGATION OF DATA IN CALCULATING ADVERSE IMPACT
by Art Gutman Ph.D., Professor, Florida Institute of Technology
The case is Stagi v. AMTRAK (2010 U.S. App. Lexis 17261, 8/16/10) in which a class of women alleged that a policy of requiring union employees to have one year of service in a current position before being promoted to a management position adversely impacted previously laid off women, a policy termed a “one-year blocking rule.” The case was filed by two named plaintiffs, both of whom were in management positions when their jobs were eliminated. Both women applied for management positions in the year following their layoff, and both were prevented from doing so and had to accept “bump down” positions based on their seniority. The case itself features a battle of experts. The plaintiff’s expert (Dr. Killingworth) found adverse impact when he aggregated data from 716 feeder pools into a large aggregated pool, whereas AMTRAK’s expert (Dr. Griffin) found no adverse impact by analyzing each individual feeder pool. The District Court for the Eastern District of Pennsylvania granted summary judgment for AMTRAK, ruling that the plaintiff’s evidence of adverse impact lacked both statistical and practical significance (U.S. Dist. LEXIS 71207, 8/12/09). In reversing the district court, the 3rd Circuit ruled:
Although it was a close case, the district court should not have granted the employer’s motion for summary judgment. The employees’ expert’s decision to aggregate the data, although not obviously correct, was also not obviously incorrect, and so there remained a genuine issue of material fact–whether the one-year rule caused a disparate impact on the employer’s female employees.
The 3rd Circuit cited “good reasons” for aggregating the data as opposed to “picking and choosing a model which will generate the most favorable results for the plaintiffs’ case” and that there is “no compelling reason” to not use aggregated data. The 3rd Circuit also noted that increased numbers makes it more likely to exclude chance as a cause of adverse impact. The 3rd Circuit cited other cases in which compelling arguments for aggregating data were made (Lilly v. Harris-Teeter Supermarket (CA4, 1983), Eldredge v. Carpenters (CA9 1987), Cook v. Boorstin (CA DC 1985) & Capaci v. Katz & Besthoff (CA5 1983). Obviously, these are older cases. Thus, we should stay tuned to other cases on this issue as they emerge.